In this paper we present a scheme able to determine malicious node, using network parameters. The contribution of this paper includes fine grained analysis (FGA).We study the malicious node and reduces packet losses in MANETs. The most existing MANETs consider packet loss. To detect truly malicious node and analyses innocent nodes .There may be several reasons for packet losses, such as interference, queue overflow, node mobility. Our proposed model achieves in false positive rate and increase in the rate of detection. Many of these approach that to be identifying the reason of packet losses can react accordingly. Mobile ad hoc networks have gained much importance due to its simplicity and low cost of deployment. The decentralized management of this network makes it susceptible to various attacks. Packet dropping is one of the major security issues. Mobile ad hoc networks are a kind of ad hoc wireless networks in which nodes are mobile. The paper concludes with the necessity of motivating a malicious node in case of malicious packet dropping in wireless ad hoc networks. Using these protocols such as DSR and AODV. In general, routing protocols for MANETs designed for predict the correct path to deliver the packets. These trust based security mechanisms for MANETs to further evaluate its effectiveness.